I believe that most complex business problems have a simple yet elegant solution waiting to be discovered. I enjoy using data as a powerful tool to uncover these solutions, transforming any industry in which it is applied.
"Without data you’re just another person with an opinion."
— W. Edwards Deming
A seasoned Data Scientist at CVS Health, working with the Pharmacy Analytics and Insights team, and a graduate degree in Business Analytics from the University of Illinois Chicago. My area of focus lies in forecasting, predictive modeling, distributed machine learning, and advanced analytics.
IoT, smart phones, fitness trackers and other technologies are generating tremendous amounts of multi-structured data that needs to be leveraged with robust analytics to transform the way businesses and industries are making important decisions and utilizing their resources.
I intend on helping organizations to capitalize on data technologies and BI to in turn make informed and powerful decisions.
As part of the Analytics & Behavior Change vertical, I leverage cutting-edge technologies and tools to develop scalable, efficient machine learning and optimization pipelines. These solutions address challenges at the scale of CVS’s retail pharmacy supply chain, driving significant business value and enhancing the experience for millions of customers nationwide.
Prior to this, I had over 5+ years of experience working at a leading software & consulting company, for one of the largest telecom clients in Austria. I designed & developed s/w products, right from ideation to deployment, with the focus on helping businesses and improving people's lives.
During my time at client’s headquarters in Vienna, I had the opportunity to analyze, understand, and extract data to derive meaningful insights that have a positive impact in defining customer’s strategy and progress.
Linkedin Download My Resume
Retailers like Walmart face difficulties in developing timely sales projections at a granularity that allows them to make exact inventory adjustments. We address this by leveraging distributed programming in Spark and Facebook’s Prophet library to build hundreds of models and forecasts for each store-department combination, thus drastically lowering the time to train the entire collection of time-series models.
Know more
Millions of records are created and downloaded every day, but only a select few are considered popular. Identifying popular music records helps streaming platforms make an efficient feedback loop. The project investigates various acoustic features to crack the secret sauce behind popular songs and use this to develop prediction & recommendation models. On top of that, it also involves visualizing weekly charts data to answer riveting questions related to top-ranking artists and songs.
Know moreFor P2P online loan platforms, lending loans to ‘risky’ applicants is one of the largest sources of financial (credit) loss. We use various EDA techniques to identify driving factors that indicate if a person is likely to default and utilise this knowledge for their portfolio & risk assessment. To improve borrower scoring quality and enable lenders to devise investment strategies, we developed data-driven models that predict the probability of default and estimate annual returns.
Know more
Consumers frequently rely on hotel reviews to make booking decisions. Managerial responses to such reviews may strengthen existing customer loyalty and convert dissatisfied consumers into loyal ones, but these responses are poorly understood. The project provides a framework using a deep-learning approach (BERT) to define the criticality of reviews and prioritize responses. In addition, we also analyse if the hotel managers are correctly addressing customers' concerns based on semantic similarity between review-response.
Know moreEvery year, approximately 15 million students apply to a university to pursue higher education. The project focuses on optimizing the university selection process by predicting the chance of admission and suggests top recommendations based on the student’s profile (like competitive exam score, years of work experience, etc.). In order to make this tool publicly accessible to students, we deploy the Flask app with an ML model using AWS Elastic Beanstalk.
Know more
Online reviews, both good and bad, have an impact on business. Keeping an eye on what customers say about business is crucial to understand what’s working well and what needs improving. We look at Yelp restaurant reviews and apply vectorization & sentiment analysis to reveal both positive and negative aspects of customers’ experience. NLP comes in handy to sift through hundreds of reviews and get the actionable insights needed to point business in the right direction.
Know moreUnlike the traditional way of segmenting markets based on demographics, we apply clustering techniques on purchasing behaviour (volume, frequency, susceptibility to discounts, and brand loyalty). This enables clients to design more cost-effective promotions that are precisely targeted to specific needs of consumer. In a long run, it benefits the company by efficiently using their corporate resources and marketing budget to make better strategic decisions.
Know more
Shubham joined my team when I was taking care of one of the most important sales applications within our company as Delivery Lead. From day one he was one of the most reliable, versatile go-to guys in our team. Whatever challenge we faced he was always quick with offering a solution and he was also a very big help in syncing up with our offshore team in India. Always ridiculously motivated, ready for action, eager to learn and hungry for more responsibility he soon started to take over Business Analyst-duties as well (while at the same time still providing his much needed developer-skills and saving the day on multiple occasions). An exceptionally talented colleague who earned my highest recommendation and I can't wait to get another chance to work with him again.
Shubham is an outstanding professional, among the most promising associates we have taken on. His input is creative and insightful and I trust him to manage our more complex tasks. He has consistently demonstrated his impeccable organization abilities by getting hold of the projects from the conceptualization phase itself. Thereafter, he is extremely efficient in implementing the strategies and adapting to the ever-changing environment of the project. He also shows his proactivity by taking the initiative in constantly upgrading his knowledge and keeping himself informed about the latest in the field. Shubham embodies a perfect mix of personal impact, analytical ability, and excellent problem-solving skills - making him a true asset to any company.
Shubham was one of the colleagues that was always open to take over new challenges. He is very talented, extremely precise when analyzing topics, and can lead team members to fulfill their goals. He is a proficient developer, and equally skilled business analyst. Shubham is highly motivated and driven by an amazing passion to succeed. Even under pressure he is able to meet timelines and is always eager to help all team members. I am very proud that we worked together and I am sure he will help more teams to meet their highest potential. I really hope we once can work together again. Until then I am sure he will help a lot of clients in other projects with his expertise and his leadership mentality.
